Public Cloud Openstack – Flexible Engine
ModelArts : Your AI models on your cloud
[Features – Data Analysis]
ModelArts is the AI development platform on the Orange Business Services public cloud: Flexible Engine. It allows you to develop your AI models for all your use cases: image, video, voice, object detection, scoring, recommendations and exception detection.
On ModelArts, integrate, build, train and deploy your AI models from end to end. Integrate and label your own datasets, or through external vendors, train your models in record time using GPU VM templates, and apply your model via inference. Then deploy your model in the cloud or on the edge.
ModelArts is an inclusive platform that adapts to all levels of expertise:
- ExeML: your models without code, create your AI models through a graphical interface in three steps :
- Upload and label your data
- Train your model in the cloud
- Apply and deploy your model
- ModelArts for beginners: get pre-integrated models and train your models by inheritance or via TensorBoard. Set only output paths and logs. Hyperparameterization and compression (quantization and pruning) are done automatically and model training is done without code.
- ModelArts for developers: take ownership of your AI models from start to finish. Gain speed thanks to data pre-processing frameworks: filtering, labeling, versioning. Develop your models in python on jupyter notebooks.
A complete platform
- Cloud performance at the service of AI
The frameworks integrated into ModelArts are optimized to reduce your processing times as much as possible, without compromising the reliability of your models. Train your models in record time thanks to VM templates equipped with Nvidia V100 GPUs.
- Accessible AI
Thanks to its wide range of features, ModelArts adapts to all levels of expertise, create models without code via a graphical interface, or develop them end-to-end in python on jupyter notebooks.
- Master your AI from start to finish
CS offers geographic location analysis functions to analyze geospatial data in real time. You can perform yaw detection and geo-fencing via SQL statements.
Deploy prospect scoring using data collected from your CRM. Analyze customer flows and typologies in a sales area. Precisely identify your targets and adjust your marketing strategies.
Fraud and anomaly detection
Detect the weak signals of an anomaly in your industrial device: material deterioration, breakdown etc. Identify intrusions, documents or fraudulent behavior.
Pre-select the best candidates through intelligent resume filtering, identify your top talent, and identify typical career paths.
Offer your customers services based on artificial intelligence: home automation, shopping, video games, etc.
AI at the service of your organization
Access ModelArts from the user console of your Flexible Engine, create and deploy your models directly to the cloud.
Optimize your processing times and costs
Optimize your training times with the power of the Flexible Engine cloud and its GPU templates, which help minimize your time to market and your infrastructure costs.
Offer new innovative services
Integrate AI into your existing processes and services, or offer new ones. ModelArts is accessible to all levels of users, with or without AI expertise, to better support you in your innovation.